I was forced to replace a vintage Engenius radio bridge with a Tranzeo SL2 tonight. Since then, the log has been filled with disconnection messages, mostly for “4-way handshake timeout.” Log file is attached as a file. Hardware retries were set to 4; raised them to 6 and then 8 and got no relief. AP configuration is:
[/size]I saw this problem beaten to death [u]here[/u] with no real resolution. I am running 5.5 on my router, which is an RB433. The old Engenius had a solid radio connection with no dropouts of this type, but had to be replaced due to water damage. I am running out of radio brands to use as this bridge. Any suggestions? log.0.txt (10.1 KB)
The Tranzeo engineer suggested I change the group key update time on the AP from the default 00:05:00 to 01:00:00, which is what the Tranzeo expects. It didn’t solve the seizures, it just cut them down to one per hour. He also suggested I set the unit to simple WPA/TKIP, which I did, but that doesn’t seem to have made a difference.
I’ve gone through two Tranzeos (both HW rev 2 and 3) and swapped out the AP radio card last night with a brand new one. I’ve never seen this behavior from other Tranzeos being used as CPEs on our other towers.
Guess I jinxed myself. The same problem just started happening on my oldest access point, an RB532, which was running 5.6, upgraded today to 5.7 but it didn’t help.
[/size]
…over and over and over. And this time they are subscriber CPEs, not infrastructure bridges. I am going to have some peevish subscribers tomorrow morning.
Looks like our problem was that our AP radio card (UBNT XR2) was going sour. It appears that one symptom of this is Tranzeos at the edge of their effective range start de-authing. Replaced the card with a fresh card and problems magically disappeared.
I’m going to post a copy of this diagnosis [u]here[/u] too, as I see many others have seen this problem.